Journal list menu

Volume 586, Issue 23 p. 4215-4222
Short communication
Free Access

Novel function of transthyretin in pancreatic alpha cells

Yu Su

Yu Su

Department of Diagnostic Medicine, Graduate School of Medical Sciences, Kumamoto University Hospital, Kumamoto, Japan

Search for more papers by this author
Hirofumi Jono

Corresponding Author

Hirofumi Jono

Department of Diagnostic Medicine, Graduate School of Medical Sciences, Kumamoto University Hospital, Kumamoto, Japan

Department of Clinical Pharmaceutical Sciences, Graduate School of Pharmaceutical Sciences, Kumamoto University, Japan

Department of Pharmacy, Kumamoto University Hospital, Kumamoto, Japan

Corresponding author. Address: Department of Diagnostic Medicine, Graduate School of Medical Sciences, Kumamoto University, 1-1-1 Honjo, Kumamoto 860-8556, Japan. Fax: +81 96 373 5283.Search for more papers by this author
Yohei Misumi

Yohei Misumi

Department of Neurology, Graduate School of Medical Sciences, Kumamoto University Hospital, Kumamoto, Japan

Search for more papers by this author
Takafumi Senokuchi

Takafumi Senokuchi

Department of Metabolic Medicine, Graduate School of Medical Sciences, Kumamoto University Hospital, Kumamoto, Japan

Search for more papers by this author
Jianying Guo

Jianying Guo

Department of Diagnostic Medicine, Graduate School of Medical Sciences, Kumamoto University Hospital, Kumamoto, Japan

Search for more papers by this author
Mitsuharu Ueda

Mitsuharu Ueda

Department of Diagnostic Medicine, Graduate School of Medical Sciences, Kumamoto University Hospital, Kumamoto, Japan

Search for more papers by this author
Satoru Shinriki

Satoru Shinriki

Department of Diagnostic Medicine, Graduate School of Medical Sciences, Kumamoto University Hospital, Kumamoto, Japan

Search for more papers by this author
Masayoshi Tasaki

Masayoshi Tasaki

Department of Diagnostic Medicine, Graduate School of Medical Sciences, Kumamoto University Hospital, Kumamoto, Japan

Search for more papers by this author
Makoto Shono

Makoto Shono

Department of Diagnostic Medicine, Graduate School of Medical Sciences, Kumamoto University Hospital, Kumamoto, Japan

Search for more papers by this author
Konen Obayashi

Konen Obayashi

Department of Diagnostic Medicine, Graduate School of Medical Sciences, Kumamoto University Hospital, Kumamoto, Japan

Search for more papers by this author
Kazuya Yamagata

Kazuya Yamagata

Department of Medical Biochemistry, Graduate School of Medical Sciences, Kumamoto University Hospital, Kumamoto, Japan

Search for more papers by this author
Eiichi Araki

Eiichi Araki

Department of Metabolic Medicine, Graduate School of Medical Sciences, Kumamoto University Hospital, Kumamoto, Japan

Search for more papers by this author
Yukio Ando

Yukio Ando

Department of Diagnostic Medicine, Graduate School of Medical Sciences, Kumamoto University Hospital, Kumamoto, Japan

Search for more papers by this author
First published: 26 October 2012
Citations: 32

Abstract

Although transthyretin (TTR) is expressed in pancreatic alpha (glucagon) cells in the islets of Langerhans, the function of TTR in pancreatic alpha cells remains unknown. In this study, by using TTR knockout (TTR KO) mice, we determined the novel role of TTR in glucose homeostasis. We demonstrated that TTR KO mice evidenced impaired recovery of blood glucose and glucagon levels. Lack of TTR induced significantly lower levels of glucagon in the islets of Langerhans. These results suggest that TTR expressed in pancreatic alpha cells may play important roles in glucose homeostasis via regulating the expression of glucagon.

1 Introduction

Glucagon, a peptide hormone of 29 amino acids, is synthesized in and secreted from pancreatic alpha cells of the islets of Langerhans in response to mixed nutrient meals, administration of oral or intravenous (i.v.) amino acids, activation of the autonomic nervous system, and hypoglycemia [1, 2]. The main action of glucagon is to raise blood glucose levels in response to hypoglycemia by increasing hepatic glycogenolysis and gluconeogenesis [3]. The circulating half-life of immunoreactive glucagon in humans is estimated to be between 5 and 6 min [4]. It is well known that the balance between insulin and glucagon controls glucose homeostasis [5]. Thus, the primary role of glucagon is correcting hypoglycemia caused by malnutrition, fasting, and treatment of diabetes to achieve normoglycemia [6-9]. Insulin treatment in diabetes, especially type 1 diabetes, greatly increases the risk of hypoglycemia, with effects on a number of organs including brain, and causes critical systemic symptoms such as hypoglycemic coma [10-12]. In addition, an absent glucagon response plays a prominent role in a failure to correct hypoglycemia induced by insulin and may exacerbate critical systemic symptoms [8]. Therefore, tight control of glucagon secretion has significant implications for stable glucose homeostasis.

Transthyretin (TTR) is a functional plasma protein, which forms a tetramer composed of four identical subunits, and serves as a transport protein for thyroxine in association with retinol-binding protein [13]. TTR is synthesized predominantly by the liver, which is therefore the main source of TTR in plasma [14, 15]. It is well-documented that plasma TTR levels are reduced in inflammatory conditions and in malnutrition caused by surgery or chronic diseases [16-19]. However, TTR is expressed in considerable amounts in the choroid plexus, retinal pigment epithelium, pancreatic alpha and beta cells, although the function of TTR synthesized by these particular cells is largely unknown [20-24]. Certain studies reported a significantly lower plasma TTR level in diabetic patients than in non-diabetic subjects [25, 26]. Moreover, in islet beta cells of the pancreas, TTR promoted insulin stimulus-secretion coupling and protected against beta cell apoptosis in type 1 diabetes [27]. In addition, TTR has been shown to be normally expressed in pancreatic alpha cells and stored in secretory vesicles [24]. Moreover, it has also been shown that TTR co-localized with glucagon exactly in pancreatic alpha cells, suggesting that TTR synthesized by pancreatic alpha cells may be involved in glucose homeostasis [24]. However, the biological significances of TTR expressed in pancreatic alpha cells remain unknown.

In view of the evidences provided in previous reports, we hypothesized that TTR may play important roles in glucose homeostasis. In this study, we focused on TTR that is synthesized by pancreatic alpha cells, and we evaluated the possible role of TTR in expression and plasma levels of glucagon during glucose fluctuations.

2 Materials and methods

2.1 Animals

Wild-type (WT) and TTR knockout (TTR KO) mice in the C57BL/6 J background [28], were used in this study. Mice were adult males, each 8–10 weeks old and each weighing 20–25 g. The animals were maintained in a pathogen-free environment at the Center for Animal Resources and Development, Kumamoto University.

2.2 Insulin tolerance test (ITT)

In the ITT, performed after a 3-h fast, human regular insulin (1 IU/kg) was administered intraperitoneally (i.p.) to mice, and the blood glucose level was measured by using an Accu-Chek Inform Blood Glucose Monitoring System (Roche Diagnostics, Indianapolis, IN).

2.3 Fasting conditions and collection of samples

Plasma samples of mice for ELISA were collected at 0, 6, 12, and 24 h after fast. Samples of pancreas and liver for (qRT-PCR) were sharply excised, and then immediately frozen in liquid nitrogen.

2.4 Cells and cell culture

The cell lines PANC-1 (human pancreas epithelioid carcinoma cells), alpha TC1 clone 6 (mouse pancreatic alpha cells), and HepG2 (human hepatocellular carcinoma cells) were obtained from ATCC (Manassas, VA) and were cultured in DMEM (Invitrogen, Grand Island, NY) supplemented with 10% FBS. All cells were grown in 5% CO2 at 37 °C.

2.5 RNA isolation and qRT-PCR

Total RNA was isolated from each tissue specimen and treated cells by using TRIzol (Invitrogen, Carlsbad, CA), according to the manufacturer's protocol. Total RNA (0.5 μg) was reverse-transcribed to cDNA by using the ExScript RT reagent (Takara Bio Inc., Shiga, Japan) according to the manufacturer's instructions. Each PCR reaction was done with 2 μl of cDNA and 0.2 μM of each primer in a LightCycler System with SYBR Premix Ex Taq (Takara Bio Inc.). Agarose gel electrophoresis was also performed as previously described. The following primers were used: mouse glucagon: forward, 5-TGAATTTGAGAGGCATGCTG-3; reverse, 5-GGTTTGAATCAGCCAGTTGA-3; mouse TTR: forward, 5-CATGAATTCGCGGATGTG-3; reverse, 5-GATGGTGTAGTGGCGATGG-3; mouse β-actin: forward, 5-TGACAGGATGCAGAAGGAGA-3; reverse, 5-GCTGGAAGGTGGACAGTGAG-3; human TTR: forward, 5-CATTCTTGGCAGGATGGCTTC-3; reverse, 5-CTCCCAGGTGTCATCAGCAG-3; and human glyceraldehyde-3-phosphate dehydrogenase (GAPDH): forward, 5-GCACCGTCAAGGCTGAGAAC-3; reverse, 5-ATGGTGGTGAAGACGCCAGT-3.

2.6 Transfection and fasting in vitro

Cells were cultured in 12 well culture plates (Becton, Dickinson, Franklin Lakes, NJ) at a density of 2 × 105 cells per well, at 37 °C in a humidified atmosphere of 5% CO2 in air for 48 h. Confluent alpha TC1 clone 6 cells were transfected with two different kinds of siRNA against murine TTR, and PANC-1 cells were transfected with human TTR plasmid by using Lipofectamine 2000 (Invitrogen) according to the manufacturer's protocol. After incubation for 24 or 48 h in serum-free medium, total RNA and protein were isolated. Alpha TC1 clone 6 cells or PANC-1 cells transiently transfected with control siRNA or control vector were used as controls. Chemical siRNA sequences for TTR siRNA1 were as follows: TTR siRNA1 sense strand 5′-CAGUGUUCUUGCUCUAUAATT-3′, TTR siRNA1 antisense strand 5′-UUAUAGAGCAAGAACACUGTT-3′ (Sigma–Aldrich, Tokyo, Japan). For evaluation of the effect of fasting on cells, after cells were incubated in serum-free high-glucose or low-glucose DMEM for 72 h, and thereafter total RNA was isolated.

2.7 Protein extraction form pancreas and pancreatic islets

Protein in the total pancreas and isolated pancreatic islets from mice was obtained by using acid ethanol (15% 1 M HCl, 75% ethanol, 10% H2O).

2.8 Isolation of pancreatic islets

Mouse pancreatic islets were isolated by means of collagenase digestion. Mice were anesthetized by i.p. with sodium thiopental. Collagenase (collagenase type S-1, 0.6 mg/ml; Nitta Gelatin Inc., Osaka, Japan) was dissolved in Hanks’ Balanced Salt Solutions (Sigma–Aldrich) with 800 KIU/ml aprotinin (Wako, Osaka, Japan). The collagenase solution was injected into the common bile duct. Pancreas were dissected and incubated in the collagenase solution at 37 °C for 20 min with shaking. These samples were then mixed with ice-cold isotonic sucrose buffer and were chilled on ice for 20 min. Clarified islets were collected for use in experiments.

2.9 Analysis of glucagon secretion

After being incubated in DMEM for 24 h, 5 groups, for each group contained 15 islets, were incubated for 1 h in Ca2+-containing HEPES-added Krebs–Ringer bicarbonate buffer (HKRB) with 2.2 mM glucose in 5% CO2 at 37 °C, followed by test incubation for 1 h in HKRB with 2.2 or 22 mM glucose.

2.10 Immunohistochemical staining

Paraffin-embedded 4-μm-thick sections were prepared and deparaffinized in xylene and rehydrated in graded alcohols. Slides were treated with periodic acid for 10 min at room temperature, after which they were incubated in 5% normal serum for 1 h at room temperature in a moist chamber. For immunohistochemical staining of glucagon, a polyclonal rabbit anti-glucagon antibody (Cell Signaling Technology, Danvers, MA), diluted 1:100 in dilution buffer, served as the primary antibody. Rabbit anti-mouse TTR antiserum, diluted 1:50 in dilution buffer, served as the primary antibody for mouse TTR. The secondary antibody was an HRP-conjugated goat anti-rabbit immunoglobulin antibody (Dako, Glostrup, Denmark) diluted 1:100 in buffer. The dilution buffer was 0.5% Bovine serum albumin (BSA). Reactivity was visualized via the DAB Liquid System (Dako), according to the manufacturer's instructions. Sections were counterstained with hematoxylin.

2.11 ELISA

The glucagon concentration in plasma, extracts of total pancreas and islets from mice, and cell lines was measured by using the Rat Glucagon ELISA Kit (Wako), according to the manufacturer's instructions for undiluted plasma samples, 1000 times dilution of extracts of pancreas and islets, 500 times dilution of cell supernatants, and 50 times dilution of cell lysates.

2.12 Statistics

Data were expressed as means ± S.D. and ±S.E.M., according to the previous study [29]. Controls and treated groups were compared by using Student's t test. The accepted level of significance was P < 0.05.

3 Results

3.1 Change in blood glucose levels in WT and TTR KO mice during the ITT

RT-PCR confirmed expression of TTR mRNA in the pancreas of mice (Fig. 1 A). In sections of pancreatic tissues, TTR was expressed in all islets from WT mice, with cells showing a preferential peripheral distribution, but islets from TTR KO mice evidenced no TTR expression (Fig. 1A). ITT showed that the blood glucose level in WT mice started to recover at 30 min after the insulin injection (Fig. 1B). However, those in TTR KO mice decreased sharply by 60 min after the injection compared with the baseline of the blood glucose level. Furthermore, TTR KO mice had continuously suppressed blood glucose levels compared with WT mice at 60, 90, 120, and 180 min after the insulin injection (Fig. 1B). We assumed that the impaired plasma glucagon levels caused the sharp decrease of the blood glucose levels in the ITT.

figure image
TTR expression in pancreatic tissues of mice and changes in blood glucose levels in WT and TTR KO mice during the ITT. (A) Expression of TTR mRNA in the pancreas of mice (top panel); TTR expression in islets from WT mice, with cells showing a preferential peripheral distribution (middle panel), but no TTR expression in TTR KO mice (bottom panel). Scale bars = 100 μm. (B) Mice received human regular insulin (1 IU/kg i.p.), and the serum glucose level was monitored at 0, 30, 60, 90, 120, and 180 min (mean ± S.E.M.). n = 6. P < 0.05.

3.2 Plasma glucagon levels in WT and TTR KO mice in the ITT and during fasting

We next investigated plasma glucagon levels in WT and TTR KO mice. As Fig. 2 A shows, although WT mice showed higher plasma glucagon levels after the insulin injection, TTR KO mice showed a much smaller increase in plasma glucagon levels in a time-dependent manner after the injection. In addition, in sections of pancreatic tissues from WT and TTR KO mice at 20 min after the insulin injection, almost all of the islets from WT mice demonstrated strong glucagon immunoreactivity, but islets from TTR KO mice showed only weak immunoreactivity (Fig. 2B). Furthermore, we compared plasma glucagon levels between WT and TTR KO mice during chronic glucose fluctuations induced by fasting. As seen in Fig. 2C, consistent with the results of the ITT, TTR KO mice showed significantly lower plasma glucagon level by fasting. In contrast to the glucagon levels, no significant difference of plasma insulin levels between WT and TTR KO mice was observed (Fig. S1).

figure image
Plasma glucagon levels in WT and TTR KO mice in the ITT and during fasting. (A) Mice underwent the ITT, and plasma glucagon was measured at 0, 10, 20, and 30 min by using ELISA (mean ± S.E.M.). n = 10. (B) Examples of pancreatic islets from mice after the ITT. At 20 min after ITT, immunohistochemistry revealed high-intensity glucagon staining in pancreatic islets from WT mice (left) and low-intensity glucagon staining in pancreatic islets from TTR KO mice (right). Scale bars = 100 μm. (C) Mice were fasted for 24 h, and plasma glucagon was measured by ELISA at 0, 6, 12, and 24 h (mean ± S.E.M.). n = 11. P < 0.05.

3.3 Glucagon content in pancreatic islets from WT and TTR KO mice

We next determined the glucagon content in pancreatic islets from WT and TTR KO mice. Under normal conditions, pancreatic islets from WT mice demonstrated a strong positive immunoreactivity for glucagon, with a preferential distribution in the periphery (Fig. 3 A). Pancreatic islets from TTR KO mice, however, showed a much weaker positive reaction (Fig. 3A). The quantitative analysis showed that in TTR KO mice, the glucagon content in the total pancreas was significantly lower than that for WT mice, i.e., about 10% of the glucagon content of WT mice (Fig. 3B). To confirm these results, pure pancreatic islets were isolated and analyzed. As seen in Fig. 3C, similar to the results for the total pancreas, the glucagon content in islets isolated from TTR KO mice was about 50% of that in from WT mice.

figure image
Glucagon content in pancreatic islets from WT and TTR KO mice. (A) An example of high-intensity glucagon staining in pancreatic islets from WT mice (top) and low-intensity glucagon staining in pancreatic islets from TTR KO mice (bottom). Scale bars = 100 μm. (B) Glucagon content in the total pancreas from WT and TTR KO mice as measured by ELISA (mean ± S.D.). n = 6. (C) Glucagon content in isolated pancreatic islets from WT and TTR KO mice, as measured by ELISA (mean ± S.D.). n = 3. P < 0.05.

3.4 Glucagon secretion from pancreatic islets isolated from TTR KO mice

To confirm the results obtained from in vivo experiments, glucagon secretion from pure pancreatic islets-isolated from WT and TTR KO mice were analyzed. Transfer from high-glucose condition to low-glucose condition has been shown to induce glucagon secretion from pancreatic islets. In our study, performed under similar conditions, pancreatic islets from WT mice showed increased glucagon secretion in the low-glucose compared with the high-glucose condition. In addition, islets from TTR KO mice had glucagon levels similar to those of WT mice in the high-glucose condition. However, no increase in glucagon level was observed in the islets from TTR KO mice in response to the low-glucose condition, in contrast to WT mice. The glucagon levels in the islets from TTR KO mice were significantly lower than those from WT mice under low-glucose conditions (Fig. 4 A). We also evaluated the glucagon contents remaining in pancreatic islets after glucagon secretion. Pancreatic islets from TTR KO mice showed a significantly lower level of glucagon than islets from WT mice-only about 40% of the WT level (Fig. 4B). Because the TTR KO islets showed the lower glucagon content, we assumed that the content of glucagon was impacted by the decrease of glucagon expression.

figure image
Glucagon secretion from pancreatic islets from WT and TTR KO mice. (A) Pancreatic islets isolated from WT and TTR KO mice were incubated under high- (22 mM) or low- (2.2 mM) glucose condition. Glucagon levels were assessed by ELISA (mean ± S.E.M.). n = 5. P < 0.05. n.s., not significant. (B) Remaining glucagon contents in isolated pancreatic islets from WT and TTR KO mice after glucagon secretion after incubation in low-glucose condition, as measured by ELISA (mean ± S.E.M.). n = 5. P < 0.01.

3.5 Changes in glucagon expression as related to TTR expression

We next determined glucagon mRNA levels after 12 h of fasting. Expression of glucagon mRNA in the total pancreas from TTR KO mice was significantly lower than that from WT mice (Fig. 5 A). While, no obvious alteration was seen in insulin mRNA expression (Fig. S2). In addition, downregulation of TTR expression by two different sequences of TTR siRNA led to a significant reduction in glucagon expression in alpha TC1 clone 6 cells at both mRNA andprotein levels (Fig. 5B, Figs. S3 and S4). In contrast, overexpression of WT-TTR by using the TTR plasmid significantly increased glucagon mRNA expression in PANC-1 cells (Fig. 5C).

figure image
Effects of TTR on glucagon expression. (A) Mice were fasted for 12 h, and glucagon mRNA expression was measured by means of qRT-PCR (mean ± S.D.). n = 14. (B) effect of blocking TTR signaling on glucagon mRNA expression in alpha TC1 clone 6 cells. Blocking was achieved by using TTR siRNA (100 nM) (mean ± S.D.). n = 3. (C) Effect of TTR overexpression on glucagon mRNA expression in PANC-1 cells. Overexpression was achieved by using TTR plasmid (0.3 μg) (mean ± S.D.). n = 3. P < 0.05; P < 0.01.

3.6 Starvation-induced changes in pancreatic TTR expression

Previous reports demonstrated that TTR levels decreased during progressive malnutrition and that this effect was linked to expression of TTR in the liver, the main organ of production of TTR circulating in the blood [17]. In our study, as expected, expression of TTR in the liver from WT mice significantly decreased. In contrast, expression of TTR in the total pancreas from WT mice markedly increased, about 2.5 times, during starvation induced by fasting for 24 h (Fig. 6a and b ). Similar to the results obtained from in vivo experiments, HepG2 cells, a liver cell line, showed a marked reduction in TTR expression after incubation under low-glucose conditions for 72 h. However, alpha TC1 clone 6 cells evidenced significantly increased glucagon mRNA expression after incubation in glucose-poor medium for 48 h (Fig. 6c and d).

figure image
Effects of starvation on TTR expression in pancreas and liver. WT mice were fasted for 24 h, and TTR mRNA expression in pancreas (A) and liver (B) was measured by means of qRT-PCR (mean ± S.D.). (C) and (D) alpha TC1 clone 6 (C) or HepG2 (D) cells were incubated under high- or low-glucose condition for 72 h. TTR mRNA expression was measured by means of qRT-PCR (mean ± S.D.). n = 3. P < 0.05; P < 0.01.

4 Discussion

A previous report suggests that, on the basis of electron microscopic evidence, TTR in pancreatic islets is mainly expressed in pancreatic alpha cells and is stored in secretory vesicles [24]. In the present study, we demonstrated that TTR KO mice, compared with WT mice, evidenced impaired blood glucose recovery and plasma glucagon levels during both acute and chronic glucose fluctuations. These results were confirmed by using isolated pancreatic islets from WT and TTR KO mice. These interesting phenomena suggest that TTR plays important roles in glucose homeostasis during glucose fluctuations, especially periods of low blood glucose levels, by regulating the amount of glucagon secreted.

One interesting finding of this study is that the lack of TTR reduced the plasma glucagon level during both acute and chronic glucose fluctuations. Glucagon and insulin constitute part of a feedback system that keeps blood glucose levels stable. The reduced glucagon level destroys the balance between insulin and glucagon, and affects the stability of blood glucose levels. These effects may be amplified by injections of insulin used to treat insulin-dependent diabetes, or by long periods of malnutrition. The pancreas releases glucagon when blood glucose levels fall too low, and glucagon facilitates the liver's conversion of stored glycogen into glucose, which is released into the bloodstream. Low plasma glucagon levels cause a failure in the relief of acute or chronic severe hypoglycemia, which affects many important organs and major physiological functions [6]. Especially, because neurons cannot use other energy sources such as fatty acids to any great degree, brain depends absolutely on glucose as a fuel [30]. Thus, rapid response and dynamic glucagon secretion when blood glucose levels are low is extremely important for maintenance of glucose homeostasis. Our data suggest that TTR performs a novel function and plays important roles in stabilizing blood glucose levels via control of glucagon. It should be noted that, despite the well-known fact that glucagon concentration is regulated by insulin concentration, no obvious difference in plasma insulin levels was observed between WT and TTR KO mice (Fig. S1). These data suggest that TTR in pancreatic alpha cells played an important role to increase the plasma glucagon levels during glucose fluctuations. A conditional TTR KO mouse which lack TTR only in pancreas may help confirming these findings. Our future studies will focus on more precise determination of the roles of TTR in glucose homeostasis.

In the present study, we found that TTR KO mice showed a significantly lower glucagon content compared with WT mice, and we found a similar phenomenon in the in vitro studies using pancreatic islets. It should be noticed that the similar numbers of pancreatic islets in whole pancreas of WT and TTR KO mice were observed by microscope (average number of WT: 32 ± 6/field; TTR KO: 28 ± 4/field). Moreover, we found that glucagon mRNA expression in pancreatic cell lines had a high positive association with the level of TTR expression. These results suggest that the lack of TTR impaired the accumulation or storage of glucagon in pancreatic alpha cells, and we considered that the reduced contents of glucagon could be caused by the decrease of mRNA levels. As shown in 1, 2, by ITT, plasma glucagon levels 20 min post-injection were lower in TTR KO than WT, and significant differences were noted from 60 min in blood glucose level. These results suggest that because pancreatic alpha cells rapidly secrete abundant glucagon in response to hypoglycemia, and since the half-life of glucagon is relatively short, 5–6 min [4], the lack of TTR may reduce the level of glucagon mRNA, which in turn leads to the deficiency of plasma glucagon levels for insulin antagonistic activity. Although the role of TTR in glycogenolysis in liver has yet to be determined, TTR may serve as one of the enhancers of glucagon expression after acute secretion of glucagon in pancreatic alpha cells. Also, we previously reported that glucose metabolism is impaired in familial amyloidotic polyneuropathy, which is caused by mutated TTR, in which amyloid deposition commonly occurs in the pancreas [31]. Previous reports demonstrated that glucagon gene expression is tightly controlled by various transcription factors, such as Pax6, Foxa1, Foxa2, and MafB/cMaf [2, 32, 33]. Of these transcription factors, Foxa2 (also called hepatic nuclear factor-3) is known to be essential for strong expression of TTR gene in the liver [34, 35]. It should be noted that Pax6 mRNA expression was significantly reduced by TTR knockdown with siRNA in alpha TC1 clone 6 cells (Fig. S5). These lines of evidence suggest that TTR in pancreatic alpha cells may also be controlled by various transcription factors, such as Pax6 and Foxa2, in response to hypoglycemia and may serve as a mediator of up-regulation of glucagon expression. In addition to the glucagon contents, because TTR is stored in secretory vesicles [24], TTR may affect the secretion of glucagon. Moreover, since we also found that the expression levels of insulin was also affected, but not obviously, by the lack of TTR (Fig. S2), further investigation is needed to explain in detail the mechanism underlying cross-talk between glucagon and TTR expression in glucose homeostasis. In addition, instead of normal cells, we used alpha TC1 clone6 and PANC-1 cell lines to obtain supportive evidence (5, 6). Unlike normal pancreatic alpha cells, alpha TC1 clone 6 cells express the high levels of glucagon by the control of the rat preproglucagon promoter and terminally differentiated. PANC-1 is an epithelioid carcinoma cell line derived from human pancreas and secret the low levels of hormones. Our future studies will focus on more precise determination by using the normal pancreatic alpha cells of the roles of TTR in glucose.

Other interesting findings of this study are that, different from the situation in the liver, hypoglycemia that was induced by insulin injections and fasting significantly enhanced TTR expression in pancreatic alpha cells, and starvation produced a similar result. These findings suggest that TTR expression may be controlled by specific transcriptional regulation in pancreatic alpha cells as described above, which differ from hepatic cells. We will clarify the mechanism in the future study.

In conclusion, unlike liver that synthesize TTR, the pancreas expresses TTR by means of a distinct mechanism. In addition to the well-known function of TTR, such as transporting thyroxine in association with retinol-binding protein, TTR that is expressed in pancreatic alpha cells may play important roles in glucose homeostasis during glucose fluctuations by regulating glucagon expression.

Acknowledgments

The authors thank Hiroko Katsura for technical assistance. This work was supported in part by the Advanced Education Program for Integrated Clinical, Basic and Social Medicine, Graduate School of Medical Sciences, Kumamoto University (Program for Enhancing Systematic Education in Graduate Schools, MEXT, Japan). The authors’ work was supported by grants from the Amyloidosis Research Committee; the Pathogenesis, Therapy of Hereditary Neuropathy Research Committee; the Surveys and Research on Specific Diseases from the Ministry of Health and Welfare of Japan; and Research for the Future Program Grant and Grants-in-Aid for Scientific Research (B) 20253742 from the Ministry of Education, Science, Sports and Culture of Japan. This work was also supported in part by the scholarship for the Graduate School of Medical Sciences, Kumamoto University, Japan.

    Appendix A A

    Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.febslet.2012.10.025.